Objective
An increasing proportion of infertile women are subjected to hysteroscopy. The effect of hysteroscopy on the pregnancy rate in assisted reproduction has been demonstrated to be favorable, but cervical dilation in the course of hysteroscopy may have an adverse effect on pregnancy outcome. We sought to investigate the effect of hysteroscopy on the risk of early miscarriage, preterm delivery, low birthweight, and other complications of pregnancy.
Study Design
This was a longitudinal retrospective cohort study at a university hospital. Data of 654 first-time singleton pregnancies between January 1997 and March 2011 in women with primary infertility were retrieved from a prospective data collection. Four cohorts were constructed based on exposure to hysteroscopy and pregnancy outcome (early miscarriage vs live birth). The primary endpoint was the duration of pregnancy at 37 weeks. Pregnancy outcomes of 167 infertile patients exposed to cervical dilation and hysteroscopy were compared with those of 327 infertile women unexposed to hysteroscopy.
Results
The incidence of miscarriage, preterm birth, placenta previa, and premature rupture of membranes after maternal exposure to hysteroscopy was similar to that in women not exposed.
Conclusion
Prior hysteroscopy in infertile women does not affect subsequent pregnancy outcome.
Both infertility and assisted reproduction technology (ART) are independent risk factors for complications during pregnancy. Preterm birth is the leading cause for low neonatal birthweight and poor infant outcome after ART. Because multiple pregnancies after ART have the greatest impact on the rate of preterm births and adverse neonatal outcome, the policy of elective single embryo transfer is increasingly being adopted to raise the proportion of singleton pregnancies. But even after adjusting for confounders singleton, pregnancies after ART come out worse than naturally conceived pregnancies. Maternal age, infertility, and ART as a procedure have been identified to be associated with a substantial adverse effect on pregnancy outcome.
Hysteroscopy (HSC) has been demonstrated to have a favorable impact on the pregnancy rates in ART, although strong evidence may still be lacking. HSC helps to identify endometrial pathology such as polyps, previously shown to interfere with embryo implantation. The embryo transfer may also be facilitated by the preceding cervical dilation. An increasing proportion of infertile women will now be subjected to cervical dilation and HSC for assessment of the uterine cavity. Hysteroscopy, especially diagnostic hysteroscopy, is considered to be a safe procedure. However, uterine perforation is the most frequent complication, and approximately half of the perforations are entry related. If the mechanical and antibacterial properties of the cervix are anatomically or functionally impaired by surgery or trauma, the cervix may become incompetent during pregnancy.
The process of cervical ripening in cervical incompetence and the mechanisms involved during preterm labor are still incompletely understood. A recent metaanalysis has shown that tissue injury caused by dilation of the cervix may be a risk factor for cervical incompetence and preterm birth in the following pregnancy. In an Australian case-control study very preterm birth between 20 and 31 weeks of gestation was significantly associated with a preceding intracervical procedure. Both curettage for abortion and treatments of precancerous changes were significantly associated with very preterm birth. To our knowledge there is to date no publication examining a possible negative impact of uncomplicated diagnostic or operative hysteroscopy during routine infertility workup on subsequent pregnancy.
The purpose of this retrospective cohort study was to investigate the effect of uncomplicated HSC on subsequent pregnancy focusing on the risk of early miscarriage, preterm delivery, and premature rupture of membranes, low birthweight, or other complications during pregnancy.
Materials and Methods
Study design/setting
We conducted a retrospective cohort study including all previously nulligravid infertile women who were treated at the Clinic of Gynecological Endocrinology and Reproductive Medicine of the University Hospital of Basel (Switzerland) and who conceived a singleton pregnancy either naturally or after various infertility treatments between January 1997 and March 2011.
Infertility was defined as the failure to conceive after a 1 year period of unprotected sexual intercourse. Natural conception was defined as conceiving a pregnancy using neither hormonal methods of ovulation induction nor assisted reproductive techniques. Any pregnancy achieved by in vitro fertilization (IVF), intrauterine insemination (IUI), intracytoplasmic sperm injection (ICSI), or oocyte donation (OD) was labeled as after-assisted conception. Live birth was defined as the birth of a living fetus, regardless of the duration of gestation.
Detailed information on personal and family history together with all information on the menstrual cycle and previous pregnancies and surgery are collected in a prospective relational web-based database (FertiMed, Reinach, Switzerland). Data generated during diagnostic workup, treatments, and subsequent pregnancies are constantly added to the database on a daily basis. Swiss legislation requires tracking of pregnancies, deliveries, and neonatal outcome if the pregnancy was conceived after IVF or ICSI and to forward key data within 12 months after childbirth to a national registry.
In addition, all pregnancies and their outcomes known to occur after natural conception and after IUI or OD are equally entered into the database. For this purpose, we routinely request a birth report from the referring obstetrician, from which we retrieve all pregnancy-related information and the details on neonatal outcome, which are then keyed into the database. The data collection methods and the study design were approved by the local institutional review board.
Participants
All infertile women conceiving a first-time singleton pregnancy naturally or with the help of assisted conception including IUI were considered eligible for this study. All participants had undergone a thorough diagnostic workup including conventional semen analysis, assessment of the body mass index (BMI) and ovarian reserve, transvaginal ultrasound in the early follicular phase and at midcycle, and screening for sexually transmitted diseases and bacterial infection of the cervix as well as hysterosalpingography if indicated.
In the case of suspected endometrial pathology, such as endometrial polyps, submucous fibroids, uterine septa, or intrauterine synechiae, a saline infusion sonography was first carried out after antibiotic treatment of any detected cervical infection with Chlamydia trachomatis , Ureaplasma urealyticum , or Mycoplasma hominis . If saline infusion sonography produced an abnormal result, the patient was referred to a reproductive surgeon for operative HSC during the follicular phase of the cycle (cycle days 5-12). Operative HSC was usually performed after cervical dilation to Hegar 9 with a 26 french resectoscope.
Confirmed eligible cases who met the study inclusion criteria were assigned to the exposed group if they had undergone HSC before achieving pregnancy and to the unexposed group if they had not undergone HSC.
Exclusion criteria
Exclusion criteria were a history of prior surgery of the uterus and/or cervix, uterine malformation (eg, uterus bicornis unicollis), ectopic pregnancy, and pregnancies that were electively terminated. Cases that were misclassified as nulligravid but had a previous pregnancy according to the birth report were excluded, and so were pregnancies after spontaneous conception with missing data on the duration of gestation.
Variables
The primary endpoint of our study was preterm birth, defined by the duration of gestation less than 37 weeks. From the FertiMed database, we extracted the following maternal and neonatal data: age at conception, BMI (expressed in kilograms per square meter) assessed during an infertility workup, smoking habits expressed as number of cigarettes consumed per day, duration of infertility in months, cause of infertility, exposure to HSC, mode of conception, duration of gestation in days, complications during pregnancy, neonatal birthweight, neonatal length, and sex of the newborn.
Potential confounders
Factors associated with poor pregnancy outcome such as maternal age, BMI, smoking habits, duration, and reason for infertility were the confounders used for multivariable analyses.
Methods to exclude or reduce bias
Because this study was carried out in an exclusively infertile population, the results may not be generalized to fertile populations. The risk of preterm birth after HSC was compared with an unexposed group within the same cohort; therefore, we consider the baseline risk for preterm birth evenly balanced in both groups. To minimize selection bias, exposed and unexposed cases were selected without knowing the outcome. All records were retrieved from the same database to prevent information bias. If the dataset was not conclusive, the missing information was retrieved from the medical file. If the data on the primary outcome was incomplete, the case was not eligible for the study. The number of patients lost to follow-up who eventually conceived spontaneously or with ART at another clinic cannot be determined in this setting.
Study size
Based on the literature, the preterm birth rate in singleton pregnancies after assisted conception is 11.4% compared with matched controls that showed a preterm birth rate of 6.1%. The relative risk at the 95% confidence level was 2.04 for preterm birth after assisted reproduction. Assuming that a relative risk of 3 or higher would reveal a clinically significant association between exposure to hysteroscopy and preterm birth, we calculated that at least 138 births were required in the group exposed to HSC to detect a 3-fold increase in the relative risk of preterm birth at less than 37 weeks of gestation at the 95% confidence level.
Quantitative variables
Causes of infertility were subdivided into polycystic ovary syndrome, combined infertility (more than 1 reason for infertility), male, tubal, none (infertility reason was removed such as resection of a septum or a polyp), unexplained, other (eg, endometriosis), and OD. Methods of conception included IVF, IUI, ICSI, spontaneous conception, and OD. Pregnancy complications were classified as early miscarriage (occurring before 12 +0 or fewer weeks of gestation), late miscarriage (defined as intrauterine death of the fetus 12 +0 or more weeks of gestation), gestational diabetes mellitus, intrauterine growth restriction, placenta previa, preeclampsia, premature rupture of the membranes, and uneventful when the birth certificate stated no pregnancy complications.
Statistical methods
The primary outcome was preterm birth before 37 weeks’ gestation.
Differences between the study groups were assessed with a χ 2 test for categorical data and Student t test for continuous variables. Continuous data were summarized showing the mean and SD. Categorical data were summarized showing the counts and percentages. To study the effect of HSC on the duration of gestation, linear regression models were performed providing differences of means with 95% confidence intervals (CIs) and corresponding P values.
In a first step, HSC alone was included in the regression model. In a second step, variables potentially influencing the outcome were also included in the regression model (covariates). Hence, the effect of HSC was adjusted for these parameters. Additionally, a regression with 1:1 propensity score matching was performed to balance the influence of the covariates. Details are described by Stampf et al.
To predict several outcomes (early miscarriage, preterm birth before 37 and 32 weeks, low birthweight, and pregnancy complications) after exposure to HSC, a logistic regression was performed. The results are presented as odds ratios (ORs) with the 95% CI and corresponding P values. The ORs were calculated unadjusted and adjusted for covariates.
Missing data for all variables were marked enabling analysis if there were no data available for a particular variable. The case was excluded if the date of birth of the child was missing. A value of P < .05 was considered significant. All analyses were done using R version 2.15.1.
Results
Participants
A total of 772 cases with first-time singleton pregnancies between January 1997 and March 2011 were retrieved from the database, and from these, 118 cases (15.3%) were excluded for various reasons ( Figure ).
The final number of cases analyzed was 654, and these were divided into 4 groups based on exposure to HSC and pregnancy outcome. Among all women achieving pregnancy, a total of 494 live births were observed.
The baseline characteristics of women exposed to and unexposed to HSC prior to conception and achieving a live birth are summarized in Table 1 and were different in several regards. Women exposed to HSC were significantly older and had longer duration of infertility, and the reasons for infertility were distributed unevenly between both groups. There were no differences with respect to the method of conception, neonatal birthweight, neonatal length, and complications during pregnancy.
| Variable | Hysteroscopy | No hysteroscopy | P value |
|---|---|---|---|
| Number of participants, n | 167 | 327 | |
| Age at conception, y a | 34.0 ± 4.64 | 32.9 ± 4.13 | .011 b |
| BMI, kg/m 2 a | 23.0 ± 4.18 | 22.4 ± 3.58 | .151 |
| Average cigarettes per day, n a | 2.1 ± 5.41 | 2.11 ± 5.51 | .983 |
| Duration of infertility, mo a | 53.3 ± 36.6 | 45.4 ± 31.7 | .018 c |
| Infertility diagnosis, % | |||
| None | 47 (28.1) | 85 (26.0) | .687 |
| Combined | 27 (16.2) | 50 (15.3) | .902 |
| Male | 9 (5.39) | 70 (21.4) | < .001 |
| Ovarian Insufficiency | 2 (1.2) | 2 (0.61) | .607 |
| Endometriosis | 43 (25.7) | 38 (11.6) | < .001 |
| PCO | 10 (5.99) | 33 (10.1) | .173 |
| Tubal | 20 (12.0) | 13 (3.98) | .001 |
| Unexplained | 9 (5.39) | 36 (11.0) | .059 |
| Method of conception, % | |||
| Natural | 69 (41.3) | 138 (42.2) | .486 |
| ICSI | 59 (35.3) | 109 (33.3) | .927 |
| IVF | 9 (5.39) | 27 (8.26) | .732 |
| IUI | 24 (14.4) | 48 (14.7) | .329 |
| Oocyte donation | 6 (3.59) | 5 (1.53) | 1.000 |
| Neonatal birthweight, g | 3249 | 3190 | .317 |
| Neonatal length, cm | 49.3 | 49.3 | .873 |
| Sex of the baby, % | .296 | ||
| Male | 91 (54.8) | 161 (49.4) | |
| Female | 75 (45.2) | 165 (50.6) | |
| Complications during pregnancy, % | |||
| Uneventful | 109 (65.3) | 222 (67.9) | .628 |
| Gestational diabetes | 6 (3.59) | 10 (3.06) | .961 |
| IUGR | 10 (5.99) | 11 (3.36) | .258 |
| Other | 23 (14.78) | 46 (14.07) | 1.000 |
| Placenta previa | 3 (1.80) | 3 (0.92) | .411 |
| Preeclampsia | 8 (4.79) | 12 (3.67) | .721 |
| PROM | 8 (4.79) | 23 (7.03) | .437 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
